An innovations approach to fault detection in sensor networks - A case study

Abstract

Sensor Networks are widespread across a large spectrum of applications. Some are even deployed in critical infrastructures, as the case of Structure Health Monitoring networks. However the sensors, due to its nature and to the harsh environmental conditions that they withstand, are subject to faults that may compromise the information quality that is indispensable for an effective safety control. In this report we present an Innovations Approach to Fault Detection in Sensor Networks case study. We study the possible application of the model based innovations approach to fault detection in sensor networks in a SHM context. This report gives a brief introduction to the model-based Fault Detection and Isolation (FDI) and a concise review of its state of the art. Next we introduce a data based FDI approach in the context of sensor validation. Given the necessary background we describe the method for sensor fault detection used in the case study. We then present the obtained results and final considerations.